Fetuin-A: a Novel Link Between Obesity and Related Complications

REVIEW Fetuin-A: a novel link between obesity and related complications

JF Trepanowski, J Mey and KA Varady

Fetuin-A (FetA) is a 64-kDa glycoprotein that is secreted from both the liver and adipose tissue. Circulating FetA is elevated in obesity and related disorders including type 2 diabetes mellitus, nonalcoholic fatty liver disease and the metabolic syndrome; and a FetA-related parameter, caliciprotein particle, is highly relevant to vascular calcification in overweight/obese patients with chronic kidney disease. FetA level is also associated with impaired insulin sensitivity and glucose tolerance. Accumulating evidence suggests that elevated FetA level causes impaired glycemic control, as FetA has been implicated in impairment of insulin receptor signaling, toll-like receptor 4 activation, macrophage migration and polarization, adipocyte dysfunction, hepatocyte triacylglycerol accumulation and liver inflammation and fibrosis. Weight loss, aerobic exercise, metformin and pioglitazone have each been shown to be effective for reducing FetA level.

International Journal of Obesity (2015) 39, 734–741; doi:10.1038/ijo.2014.203

INTRODUCTION relationship, epidemiologic, mechanistic (cell, animal and human Fetuin-A (FetA) is a 64-kDa glycoprotein that is found in relatively models) and interventional research are each summarized. high concentrations in human serum (300–1000 μgml− 1).1–3 Although all relevant research is discussed, special emphasis is Fetuin was first discovered in 1944 in bovine calves, and it given to studies that have been published as the most recent 19,20 derived its name from ‘fetus’ to reflect the observation that fetal reviews of FetA and obesity. serum contained the highest concentration of this protein.4 Fetuin was renamed ‘fetuin-A’ in 2000 upon the discovery of a fetuin-like SEARCH STRATEGY molecule termed ‘fetuin-B’.5 Human FetA is also known as α2- Heremans-Schmid glycoprotein (AHSG);6 this naming honors the A PubMed search was conducted through October 2014 of (independent) discoverers of the human homolog of FetA, publications in English. The following input was used for the Heremans7 along with Bürgi and Schmid.8 search: (FetA or AHSG) and (obesity or diabetes or metabolic In human adults, FetA is mainly expressed and secreted from syndrome or NAFLD or cardiovascular disease (CVD) or kidney the liver and adipose tissue. Epidemiologic research consistently disease or nephropathy or glucose or palmitate or insulin receptor observes elevated circulating FetA in obesity9 and related or inflammation or toll-like receptor 4 or adipocyte or diet or complications, such as type 2 diabetes mellitus (T2DM),1 the exercise or metformin). This search yielded 599 potentially metabolic syndrome10 and nonalcoholic fatty liver disease relevant articles. Additional articles were identified using the (NAFLD).11 Moreover, FetA level is associated with many references listed in these manuscripts. Intervention studies were parameters related to metabolic health, such as insulin considered for review only if they: (1) included a subject sensitivity,3 glucose tolerance,11 circulating lipid levels10 and population that was overweight/obese (body mass index (BMI) ⩾ −2 circulating levels of both pro- and anti-inflammatory proteins.12 25.0 kg m ) and/or presented with T2DM and/or presented FetA also has an important role in calcification inhibition,13 with NAFLD; and (2) examined an intervention that typically particularly in patients with chronic kidney disease (CKD).14 reduces body weight and/or improves insulin sensitivity. Articles Recent research is the uncovering mechanisms that underlie the were selected for inclusion in this review if they provided evidence relationship between FetA and obesity-related complications. for an association between FetA level and obesity or related fi Much of this research can be categorized as explaining either metabolic conditions. A total of 87 articles were nally selected. (1) how glucolipotoxicity induces FetA expression or (2) how FetA promotes metabolic dysfunction. EPIDEMIOLOGIC DATA Intervention trials demonstrate that diet,15 aerobic exercise,16 gastric bypass surgery,9 metformin17 and pioglitazone18 can each Obesity be employed to reduce circulating FetA. Reductions in FetA level Associations between FetA level and obesity-related parameters are often associated with improvements in insulin-related para- are widely reported in the epidemiologic literature. BMI,21 visceral meters and/or circulating adiponectin.15,16 adipose tissue (VAT)1 and leptin concentration22 are each This is the first comprehensive review to exclusively focus on positively associated with FetA concentration. Furthermore, the relationship between FetA and obesity. To expound on this change in FetA level is associated with the change in visceral

Department of Kinesiology and Nutrition, University of Illinois at Chicago, Chicago, IL, USA. Correspondence: JF Trepanowski, Department of Kinesiology and Nutrition, University of Illinois at Chicago, 1919 West Taylor Street, Room 506F, Chicago, 60612, IL, USA. E-mail: [email protected] Received 19 June 2014; revised 4 October 2014; accepted 2 November 2014; accepted article preview online 3 December 2014; advance online publication, 6 January 2015 Fetuin-A in obesity and related complications JF Trepanowski et al 735 adipose tissue over a 5-year period.23 Circulating FetA is elevated pressure, fasting glucose level, fasting insulin level, the home- in insulin-resistant obesity compared with insulin-sensitive ostasis model of insulin resistance, OGTT-derived 2-h glucose level obesity.24 and OGTT-derived 2-h insulin level.3,11,37,41,42 FetA concentration The relationship between FetA and obesity appears to be has been found to be negatively associated with high-density mediated at least in part by genetics. A variant of the AHSG gene lipoprotein cholesterol level, adiponectin level and insulin that is associated with lower FetA level is more common among sensitivity measured with the euglycemic clamp.3,11 The AHSG normal weight men compared with overweight and obese men.25 gene is located on chromosome 3q27, a region that has been In addition, bi-directional Mendelian randomization shows evi- identified as a metabolic syndrome susceptibility locus.43 dence of a casual association from circulating FetA to BMI, along 26 with no evidence of reverse causality from BMI to FetA. Nonalcoholic fatty liver disease Circulating FetA is elevated in NAFLD (even when patients with Type 2 diabetes mellitus the same glycemic status are compared),11 and is positively Incident diabetes risk increases in individuals with higher FetA associated with liver fat.3 FetA concentration is also positively concentrations even after adjustment for potential mediators, associated with the rs738409 I148M variant of patatin-like including: body weight, BMI, waist circumference, blood pressure, phospholipase domain-containing 3,44 which is the major blood lipid levels, race, fasting glucose level, HbA1C level, determinant of liver fat content in the general population. C-reactive protein level, adipocytokine levels and liver enzyme A positive association between FetA level and the liver fibrosis levels.1,2,27 A recent investigation found that a positive association score index has been observed.45 However, other studies between FetA concentration and incident diabetes risk existed have reported no association between FetA level and liver only in female participants.21 However, previous research found histology.17,44,46 A negative association between FetA concentra- that this association applied similarly to both genders.1,2,28 tion and the calculated NAFLD fibrosis score was also reported,47 Gestational diabetes patients have elevated FetA level compared although this calculation does not involve measurements derived with healthy pregnant women and non-pregnant controls.29 from liver biopsy or liver imaging. Among NAFLD patients, FetA Moreover, FetA concentration is positively associated with both level has been reported to be positively associated with carotid fasting C-peptide concentration and C-peptide/blood glucose artery intima-media thickness in one investigation,46 and ratio in both gestational diabetes patients and healthy pregnant negatively associated in another.47 Differences in NAFLD diagnosis women.29 Isolated impaired glucose tolerance, but not isolated methodology (liver biopsy for the study reporting a positive impaired fasting glucose, is associated with higher FetA concen- association, abdominal ultrasonography for the study reporting a tration when compared with normal glucose tolerance.11 The negative association) may explain these discrepant findings. reason for this is currently unclear but may be related to the greater degree of peripheral insulin resistance that is typically Chronic kidney disease observed in isolated impaired glucose tolerance.30 In individuals Vascular calcification is a prominent feature of CKD and an with impaired glucose tolerance, but not normal glucose established risk factor for cardiovascular events and cardiovascular tolerance, FetA concentration is negatively associated with insulin mortality.48 FetA is a potent extraosseous calcification inhibitor,13 secretion during an oral glucose tolerance test (OGTT).31 and in studies examining CKD patients (typically with a mean BMI AHSG gene single-nucleotide polymorphisms have been asso- − of ~ 25 kg m 2), FetA level is inversely associated with calcification ciated with prevalent type 2 diabetes.32,33 Yet, a recent Mendelian scores, cardiovascular events and cardiovascular mortality (see randomization study observed that genetically predicted FetA Evrard et al.49 for a review). On the other hand, the number of level was not associated with prevalent diabetes, incident diabetes studies that have examined the relationship between FetA level or fasting glucose.34 This suggests that FetA level may not be and vascular calcification (and related parameters) in a predomi- causally related to diabetes risk. However, this study did not nately overweight/obese population is very small, and highlights measure non-esterified fatty acid (NEFA) level, which has been the need for more work in this area. Mehrotra et al.50 observed a recently shown to interact with FetA level to predict the degree of paradoxical positive association between FetA level and coronary insulin resistance.35 Future Mendelian randomization studies and artery calcification score in patients with diabetic nephropathy interventional studies should measure NEFA concentration along who were not receiving dialysis. However, the researchers also with FetA concentration to better determine whether the latter is found a positive association between FetA level and glomerular causally related to diabetes risk. − filtration rate (which was a mean of 30 ml min 1 per 1.73 m), The relationship between circulating FetA and CVD risk appears suggesting that the patients’ average renal function was ‘too to be modified by the diabetes status. Non-diabetics with higher healthy’ to observe the expected relationship between FetA level FetA level have decreased risks of incident CVD36 and CVD-related and coronary artery calcification score.50 Another limitation of this mortality;37 this is likely due to the vascular calcification inhibitory study is its measurement of total FetA concentration as a putative ability of FetA.13 On the other hand, type 2 diabetics with higher marker of extraosseous calcification stress, because research has FetA level have increased risks of incident CVD36 and CVD-related shown that measurement of total FetA concentration often fails to mortality.37 Two studies have examined the association between yield the expected results in studies examining CKD.14 Rather, the circulating FetA and prevalent peripheral arterial disease among more effective method that is now employed involves measuring type 2 diabetics, with one study reporting a positive association,38 the fraction of total circulating FetA that is incorporated into and another study reporting a negative association.39 Methodo- calciprotein particles.14 Calciprotein particles have been positively logical issues in the latter study (specifically very selective study associated with coronary artery calcification scores, aortic inclusion criteria) may explain these discrepant findings.40 stiffness and procalcific cytokine production in (predominately) overweight/obese CKD patients.14,51 Furthermore, the time to Metabolic syndrome transformation from spherical primary calciprotein particles to Relationships between FetA level and parameters related to the spindle-shaped secondary calciprotein particles (reflecting metabolic syndrome have been researched. FetA level has been serum calcification propensity)52 was significantly associated found to be positively associated with the following: total with all-cause mortality in (predominately) overweight/obese cholesterol level, low-density lipoprotein cholesterol level, non- patients with stages 3 and 4 CKD.53 high-density lipoprotein cholesterol level, triacylglycerol level, So far, we are unaware of any human studies that have waist circumference, systolic blood pressure, diastolic blood examined the relationship between circulating FetA and

© 2015 Macmillan Publishers Limited International Journal of Obesity (2015) 734 – 741 Fetuin-A in obesity and related complications JF Trepanowski et al 736 extraosseous calcification outside of the cardiovascular system. ERK 1/2 signaling pathway is involved in the FetA response to However, the recent discovery that FetA protects the kidney from endoplasmic reticulum stress.62 nephrocalcinosis in rats54 provides promise for future research in It was believed until recently that the liver was the only major humans. secretory organ of FetA. This paradigm changed with the discovery that adipocytes can also synthesize FetA. Incubation of mouse adipocytes with palmitate dose-dependently increased MECHANISTIC DATA FetA mRNA expression, protein expression and secretion; these Glucolipotoxicity induces FetA expression effects are mediated by nuclear factor-κB.63 Positive energy balance in mice has been shown to increase FetA mRNA,55 and in humans, positive energy balance was recently FetA promotes metabolic dysfunction 56 shown to elevate circulating FetA. Experiments have examined Interference with insulin receptor signaling. Srinivas et al.64 first mechanisms that underlie these phenomena, specifically examin- provided evidence two decades ago that human FetA interferes ing glucolipotoxic conditions. Figure 1 illustrates the main findings with insulin-receptor signaling at the tyrosine kinase level. Their and explains how FetA can link glucolipotoxicity to hepatic research corroborated similar reports that rat FetA65,66 and bovine steatosis. FetA67 also inhibit insulin receptor tyrosine kinase activity. Other Incubation of HepG2 cells or rat hepatocytes with palmitate work demonstrated that Fetuin-null mice were protected from stimulates nuclear factor-κB binding to the FetA promoter, age-induced68 and obesity-induced insulin resistance.69 Early thereby augmenting FetA mRNA expression, protein synthesis investigations suggested that although FetA interferes with the 57 and secretion. Palmitate-induced FetA subsequently stimulates mitogenic effects of insulin signaling, it had no effect on insulin’s triacylglycerol accumulation in hepatocytes through the metabolic actions.64,65,70,71 However, recent evidence that FetA mammalian target of rapamycin-sterol and regulatory element- blocks insulin-stimulated glucose transporter 4 translocation and 58 binding protein-1c pathway. Adiponectin inhibits palmitate- protein kinase B activation demonstrates the metabolic-interfering induced hepatic FetA expression through the adenosine effects of this protein.72 This study also clarified that FetA monophosphate-activated protein kinase pathway.58 Thus, hypo- interferes with downstream phosphorylation events in insulin adiponectinemia, which is frequently observed in obesity,59 may receptor signaling without affecting the binding of insulin to the be another cause of elevated FetA. In lean, healthy humans, a α-subunit of the receptor72 (Figure 2). trend was observed for increased circulating FetA in response to low-dose infusion of Intralipid (30 ml per hour) for 48 h.60 Inflammatory stimulation. Apart from its direct effects on the Incubation of HepG2 cells with glucose increases FetA protein insulin receptor, FetA may also promote insulin resistance by expression and enhances the AHSG gene promoter activity.61 propagating a pro-inflammatory state. In both adipocytes and These effects are mediated through the ERK 1/2 signaling monocytes, FetA treatment augments pro-inflammatory cytokine pathway.61 Endoplasmic reticulum stress appears to mediate both mRNA and protein expression while reducing adiponectin mRNA glucose- and lipid-induced elevations in FetA expression.62 The and protein expression.12,57 Furthermore, intraperitoneal bolus

APN

Cell membrane

FA FA AMPK FA G

G G ERK 1/2

P mTOR 1

Fetuin-A Lipogenic enzymes

2 Triacylglycerol accumulation SREBP-1C 3 NF-κB

Figure 1. Proposed model: glucolipotoxic-mediated upregulation of fetuin-A leads to hepatic steatosis. Excess glucose and fatty acids impose endoplasmic reticulum stress and activate ERK 1/2, which subsequently upregulates NF-κB. Reduction of NF-κB activity by AMPK is limited due to the inhibition by palmitate and under-stimulation by APN (as a consequence of fetuin-A-induced hypoadiponectinemia). Signal transduction through NF-κB promotes fetuin-A upregulation, which in turn induces mTOR phosphorylation and SREBP-1C expression. Upregulated lipogenic enzymes by SREBP-1C favor triacylglycerol accumulation. AMPK, adenosine monophosphate-activated protein kinase; APN, adiponectin; ER, endoplasmic reticulum; ERK 1/2, extracellular signal-regulated kinases 1 and 2; FA, fatty acid; G, glucose; mTOR, mechanistic target of rapamycin; NF-κB, nuclear factor-κB; SREBP-1C, sterol regulatory element-binding protein-1c.

Normal insulin signaling Fetuin-A interference

G G G G G Insulin Insulin G G G G

Fetuin-A G

GLUT4 ßß ßß P IRS1 IRS1

G PI3K PI3K G Akt Akt

GLUT4 GLUT4

Figure 2. Fetuin-A interferes with the metabolic arm of insulin receptor signaling. Left, insulin binds to the α-subunit of the insulin receptor, which initiates a cascade of phosphorylation events that ultimately promotes GLUT4 translocation and enhances glucose uptake. Right, fetuin- A binds to the ectodomain of the β-subunit. This does not affect binding of insulin to the α-subunit, but it does interfere with downstream phosphorylation events. GLUT4 translocation is inhibited, and glucose uptake is impaired. α, α-subunit of insulin receptor; β, β-subunit of insulin receptor; Akt, protein kinase B; G, glucose; GLUT4, glucose transporter type 4; IRS1, insulin receptor substrate 1; P, phosphate group (PO4); PI3K, phosphatidylinositide 3-kinase. delivery of FetA in C57BL/6 mice induces adipose tissue lower the lipid droplet size and numbers while reducing total lipid expression of IL-6 and TNF while reducing ADIPOQ mRNA and content.57 FetA treatment also lowered the protein and mRNA circulating adiponectin.12 These findings may help to explain the expressions of the adipogenic factor peroxisome proliferator- inverse association between circulating FetA and adiponectin that activated receptor-γ, as well as downstream molecules of the is frequently observed.59 peroxisome proliferator-activated receptor-γ signaling pathway, Pal et al.73 recently reported that FetA is necessary for NEFAs to including adiponectin, adipocyte protein 2 and fatty acid induce inflammation and insulin resistance via toll-like receptor 4 translocase.57 Relationships between AHSG gene single- (TLR4) signaling in both adipocytes and macrophages. The nucleotide polymorphisms and adipocyte function have also researchers also found that FetA forms a ternary complex with been reported. One study found that the rs2077119 single- NEFAs and TLR4, indicating that FetA serves as an endogenous nucleotide polymorphism had strong associations with insulin ligand for TLR4 signaling.73 Building on these findings, Stefan and inhibition of lipolysis, as well as 8-bromocyclic AMP stimulation of Häring35 indirectly assessed whether FetA mediates lipid-induced lipolysis, and weak associations with insulin stimulation of insulin resistance in humans in vivo. The researchers observed an lipogenesis and basal lipolysis.76 Another study found that the interaction effect between NEFA level and FetA level on OGTT- rs4917 single-nucleotide polymorphism was associated with derived insulin sensitivity: NEFA level was only significantly lipolytic sensitivity to terbutaline.77 associated with insulin sensitivity in individuals with high FetA levels, and FetA level was only significantly associated with insulin sensitivity in individuals with high NEFA levels.35 FetA activation of INTERVETIONAL DATA TLR4 recently was observed to promote nonalcoholic steato- Interventions that produce weight loss are effective at lowering hepatitis and liver fibrosis independently of toll-interleukin-1 FetA3,9,15–17,78,79 (Table 1). Gastric bypass surgery improves FetA receptor domain-containing adapter-inducing interferon-β.74 level, albeit not as effectively as its improvements in some Macrophage migration into adipose tissue and polarization components of the metabolic syndrome (that is, systolic blood from an anti-inflammatory M2 subtype to a pro-inflammatory M1 pressure, triacylglycerol level and fasting glucose level).9 A recent subtype are important events that promote the low-grade investigation observed that gastric bypass reduced circulating inflammatory state observed in obesity.75 Adipocyte-derived FetA FetA to a greater extent compared with sleeve gastrectomy 3 days was recently found to signal both of these events.63 Figure 3 following surgery and before improvements in homeostasis model represents a putative model to explain how FetA promotes a pro- of insulin resistance could be observed;80 the mechanism inflammatory state in adipose tissue. explaining this greater reduction awaits discovery. Aerobic exercise can reduce FetA in the absence of either weight loss or Impairment of adipocyte function. FetA may attenuate lipo- even reduction in hepatic steatosis provided that the exercise genesis and accelerate lipolysis in adipocytes, thereby promoting intensity is sufficient: whereas exercising at the anaerobic 18 obesity and insulin resistance. Treatment of mouse 3T3L1 threshold, 60% volume of oxygenpeak (ref. 81) and 60–75% adipocytes and human pre-adipocytes with FetA was found to heart rate (HR)max (ref. 82) were all unable to reduce circulating

© 2015 Macmillan Publishers Limited International Journal of Obesity (2015) 734 – 741 Fetuin-A in obesity and related complications JF Trepanowski et al 738 Blood Adipose tissue FA FA Liver-derived Adipocyte-derived FA Fetuin-A Fetuin-A Fetuin-A

Fetuin-A Fetuin-A Fetuin-A

Fetuin-A Fetuin-A Fetuin-A FA

FA FA Chemoattractant FA Adipocyte signaling FA Quiescent FA macrophage FA

M2 macrophage

Phenotype conversion Fetuin-A Pro-inflammatory M1 cytokine release macrophage

Figure 3. Proposed model: fetuin-A promotes adipose tissue inflammation. Fetuin-A originating from hepatocytes and adipocytes sends chemoattractant signals that induce macrophage infiltration into adipose tissue and subsequent conversion to a classically activated M1 subtype. Fetuin-A then presents fatty acids to the TLR4 receptors on both M1 macrophages and adipocytes, thereby propagating the release of pro-inflammatory cytokines. FA, fatty acid; TLR4, toll-like receptor 4.

FetA, exercising at 85% HRmax was able to achieve this in only 7 Unfortunately, due to the different assay techniques that have days.83 A low dose of metformin (500 or 750 mg per day for been employed in the epidemiologic literature,14 along with the 3 months) was ineffective at reducing FetA level,18 but higher many confounders existing across studies (for example, insulin dosing (2500 or 3000 mg per day for 6 months) was successful.17 sensitivity, fat mass, gender, age and so on), clinically relevant cut Pioglitazone (15 or 30 mg per day for 3 months) has also been points of FetA concentration in obesity and related disorders such observed to lower circulating FetA.18 as T2DM, NAFLD, the metabolic syndrome and CKD are currently As has been previously observed elsewhere,84 each of these unknown. A standardized assaying procedure will likely need to interventions is capable of activating adenosine monophosphate- be widely adopted before such cut points emerge. activated protein kinase, which itself has been shown to inhibit It is likely that FetA has a causal role in the development and hepatic FetA expression.58 Future research should confirm progression of obesity-related complications, because FetA has whether adenosine monophosphate-activated protein kinase been implicated in impairment of insulin receptor signaling, TLR4 activation is an important factor in intervention-induced reduction activation, macrophage migration and polarization, adipocyte of circulating FetA. Future work should also measure phosphory- dysfunction, hepatocyte triacylglycerol accumulation and liver lated FetA, as phosphorylation is necessary for insulin receptor- inflammation and fibrosis. Weight loss through diet or surgery, tyrosine kinase inhibition by this protein.85 Interestingly, a single aerobic exercise, metformin and pioglitazone have each been bout of treadmill walking that expended 350 kcal reduced serum shown to be effective for reducing FetA, provided that the phosphofetuin without affecting total circulating levels.86 Fasting intensity/dose of the intervention is sufficient. insulin level and homeostasis model of insulin resistance were also Going forward, research should investigate whether FetA has a reduced, and while the data are preliminary, they suggest that role in the progressive deterioration of β-cell function that occurs changes in phosphofetuin, and not total FetA, may be associated during the etiology of T2DM. Such a role is likely when with these improvements. considering the importance of TLR4 activation in lipid-induced Reduced FetA in response to intervention is often associated β-cell dysfunction,87 the discovery that FetA serves as an with improvement in insulin-related parameters. Indeed, reduc- endogenous ligand for TLR4, and the finding that OGTT- tions in FetA have been associated with improvements in fasting derived insulin secretion is negatively associated with FetA level insulin,9 homeostasis model of insulin resistance,9,79 OGTT-derived in individuals with impaired glucose tolerance. Future research insulin sensitivity,83 hepatic insulin resistance (fasting insulin × should additionally clarify whether adenosine monophosphate- fasting hepatic glucose production)16 and metabolic flexibility activated protein kinase activation is involved in intervention- (insulin-stimulated respiratory exchange ratio − fasting respiratory induced lowering of circulating FetA. Other areas in need of exchange ratio).16 Also, reductions in FetA have been associated clarification include the relative contributions of the liver and with increases in circulating adiponectin.15,16 adipose tissue to circulating FetA levels, the specific mechanism by which FetA promotes macrophage migration and polarization, the factors that underlie the association between CVD and CONCLUSIONS FetA level in individuals with T2DM and the ability of FetA to The evidence is clear that FetA is elevated in obesity protect against extraosseous calcification outside of the cardio- and is associated with many obesity-related complications. vascular system.

Table 1. Interventions for reducing fetuin-A

Reference Subjects Intervention Trial length ΔBody ΔFetuin-A weight (%) (%)

83 Malin et al. n = 13, M/F AEx (85% HRmax, 60 min per day, 7 days per week) 7 days — ↓11% age 51 ± 3 years BMI 33 ± 1kgm−2 16 Malin et al. n = 20, M/F AEx (85% HRmax, 60 min per day, 5 days per week) 12 weeks ↓6% ↓8% age 66 ± 1 years BMI 34 ± 1kgm− 2 Mori et al.18 n = 8, M/F AEx (anaerobic threshold, 40 min per day, 3–5 days 3 months —— age 62 ± 18 years per week) BMI 27 ± 4kgm−2 81 Schultes et al. n = 14, F AEx (60% VO2 peak, 60 min per day, 3 days per week) 6 weeks —— age 43 ± 3 years BMI 37 ± 2kgm− 2 82 Yang et al. n = 40, F AEx (60–75% HRmax, 45 min per day, 5 days per week) + REx 12 weeks —— age 45 ± 10 years (unspecified intensity, 45 min per day, 5 days per week) BMI 28 ± 2kgm− 2 Blüher et al.78 n = 322, M/F CR (unspecified degree of energy restriction) 2 years ↓4% ↓42 mg l − 1 age 52 ± 7 years BMI 31 ± 4kgm− 2 Choi et al.15 n = 38, F CR (30%) 12 weeks ↓7% ↓6% Age 56 ± 7 years BMI 27 kg m − 2 Reinehr and Roth79 n = 21, M/F AEx + CR (both unspecified) 1 year ↓0.7 (SDS-BMI) ↓10% age 11 ± 3 years SDS-BMI 2.4 ± 0.4 Stefan et al.3 n = 47, M/F AEx (unspecified intensity, totaling 3 h per week) + CR ~ 9 months ↓4% ↓27% age 44 ± 2 years (unspecified) body weight 87 ± 3kg Brix et al.9 n = 75, M/F Gastric bypass 16 months ↓35% ↓19% age 40 ± 10 years BMI 46 ± 9kgm−2 Haukeland et al.17 n = 20, M/F Metformin (2500 or 3000 mg day) 6 months ↓4% ↓40 mg l − 1 age 44 ± 2 years BMI 31 ± 1kgm− 2 Mori et al.18 n = 9, M/F Metformin (500 or 750 mg per day) 3 months —— age 62 ± 6 years BMI 29 ± 5kgm− 2 Mori et al.18 n = 10, M/F Pioglitazone (15 or 30 mg per day) 3 months ↑1 BMI ↓13% age 63 ± 10 years BMI 27 ± 4kgm− 2 Abbreviations: AEx, aerobic exercise; BMI, body mass index (kg m −2); CR, calorie restriction; F, female; HR, heart rate; M, male; REx, resistance exercise; SDS, standard deviation score; VO2, volume of consumed oxygen.

CONFLICT OF INTEREST 9 Brix JM, Stingl H, Höllerl F, Schernthaner GH, Kopp H-P, Schernthaner G. Elevated The authors declare no conflict of interest. Fetuin-A concentrations in morbid obesity decrease after dramatic weight loss. J Clin Endocrinol Metab 2010; 95: 4877–4881. 10 Ix JH, Shlipak MG, Brandenburg VM, Ali S, Ketteler M, Whooley MA. Association REFERENCES between human fetuin-A and the metabolic syndrome data from the heart and soul study. Circulation 2006; 113: 1760–1767. 1 Ix JH, Wassel CL, Kanaya AM, Vittinghoff E, Johnson KC, Koster A et al. Fetuin-A 11 Ou H-Y, Yang Y-C, Wu H-T, Wu J-S, Lu F-H, Chang C-J. Increased fetuin-A and incident diabetes mellitus in older persons. JAMA 2008; 300: 182–188. concentrations in impaired glucose tolerance with or without nonalcoholic fatty 2 Stefan N, Fritsche A, Weikert C, Boeing H, Joost H-G, Häring H-U et al. Plasma liver disease, but not impaired fasting glucose. J Clin Endocrinol Metab 2012; 97: fetuin-A levels and the risk of type 2 diabetes. Diabetes 2008; 57: 2762–2767. 4717–4723. 3 Stefan N, Hennige AM, Staiger H, Machann J, Schick F, Kröber SM et al. α2- Heremans-Schmid glycoprotein/fetuin-A is associated with insulin resistance and 12 Hennige AM, Staiger H, Wicke C, Machicao F, Fritsche A, Häring H-U et al. Fetuin-A fat accumulation in the liver in humans. Diabetes Care 2006; 29: 853–857. induces cytokine expression and suppresses adiponectin production. PLoS One 3 4 Pedersen KO. Fetuin, a new globulin isolated from serum. Nature 1944; 154: 2008; :e1765. 575–575. 13 Schäfer C, Heiss A, Schwarz A, Westenfeld R, Ketteler M, Floege J et al. The serum 5 Olivier E, Soury E, Ruminy P, Husson A, Parmentier F, Daveau M et al. Fetuin-B, protein α2–Heremans-Schmid glycoprotein/fetuin-A is a systemically acting a second member of the fetuin family in mammals. Biochem J 2000; 350:589–597. inhibitor of ectopic calcification. J Clin Invest 2003; 112: 357–366. 6 Schultze H, Heide K, Haupt H. Charakterisierung eines niedermolekularen 14 Hamano T, Matsui I, Mikami S, Tomida K, Fujii N, Imai E et al. Fetuin-mineral α 2-Mukoids aus Humanserum. Naturwissenschaften 1962; 49:15–15. complex reflects extraosseous calcification stress in CKD. J Am Soc Nephrol 2010; 7 Heremans J. Les globulines seriques du systeme gamma: leur nature et leur 21: 1998–2007. pathologie In: Arscia. Brussels, 1960. 15 Choi KM, Han KA, Ahn HJ, Lee SY, Hwang SY, Kim BH et al. The effects of caloric 8 Bürgi W, Schmid K. Preparation and properties of Zn-α2-glycoprotein of normal restriction on Fetuin‐A and cardiovascular risk factors in rats and humans: human plasma. J Biol Chem 1961; 236: 1066–1074. a randomized controlled trial. Clin Endocrinol (Oxf) 2013; 79:356–363.

© 2015 Macmillan Publishers Limited International Journal of Obesity (2015) 734 – 741 Fetuin-A in obesity and related complications JF Trepanowski et al 740 16 Malin SK, Del Rincon JP, Huang H, Kirwan JP. Exercise-Induced Lowering of 41 Xu Y, Xu M, Bi Y, Song A, Huang Y, Liu Y et al. Serum fetuin-A is correlated with Fetuin-A May Increase Hepatic Insulin Sensitivity. Med Sci Sports Exerc 2014; 46: metabolic syndrome in middle-aged and elderly Chinese. Atherosclerosis 2011; 2085–2090. 216:180–186. 17 Haukeland JW, Dahl TB, Yndestad A, Gladhaug IP, Løberg EM, Haaland T et al. 42 Chen H-Y, Chiu Y-L, Hsu S-P, Pai M-F, Lai C-F, Peng Y-S et al. Association of serum Fetuin A in nonalcoholic fatty liver disease: in vivo and in vitro studies. fetuin A with truncal obesity and dyslipidemia in non-diabetic hemodialysis Eur J Endocrinol 2012; 166: 503–510. patients. Eur J Endocrinol 2009; 160: 777–783. 18 Mori K, Emoto M, Araki T, Yokoyama H, Lee E, Teramura M et al. Effects of 43 Vionnet N, Hani EH, Dupont S, Gallina S, Francke S, Dotte S et al. Genomewide pioglitazone on serum fetuin-A levels in patients with type 2 diabetes mellitus. search for type 2 diabetes–susceptibility genes in French Whites: evidence for a Metabolism 2008; 57: 1248–1252. novel susceptibility locus for early-onset diabetes on chromosome 3q27-qter and 19 Stefan N, Häring H-U. The role of hepatokines in metabolism. Nat Rev Endocrinol independent replication of a type 2–diabetes locus on chromosome 1q21–q24. 2013; 9: 144–152. Am J Hum Genet 2000; 67: 1470–1480. 20 Rasul S, Wagner L, Kautzky-Willer A. Fetuin-A and angiopoietins in obesity and 44 Rametta R, Ruscica M, Dongiovanni P, Macchi C, Fracanzani AL, Steffani L et al. type 2 diabetes mellitus. Endocrine 2012; 42: 496–505. Hepatic steatosis and PNPLA3 I148M variant are associated serum Fetuin‐A 21 Laughlin GA, Barrett-Connor E, Cummins KM, Daniels LB, Wassel CL, Ix JH. independently of insulin resistance. Eur J Clin Invest 2014; 44:627–633. Sex-specific association of fetuin-A with type 2 diabetes in older community- 45 Yilmaz Y, Yonal O, Kurt R, Ari F, Oral AY, Celikel CA et al. Serum fetuin A/α2HS- dwelling adults: the Rancho Bernardo study. Diabetes Care 2013; 36: 1994–2000. glycoprotein levels in patients with non-alcoholic fatty liver disease: relation with 22 Vörös K, Gráf L Jr, Prohászka Z, Gráf L, Szenthe P, Kaszás E et al. Serum fetuin‐Ain liver fibrosis. Ann Clin Biochem 2010; 47:549–553. metabolic and inflammatory pathways in patients with myocardial infarction. Eur 46 Dogru T, Genc H, Tapan S, Aslan F, Ercin CN, Ors F et al. Plasma fetuin‐Ais J Clin Invest 2011; 41:703–709. associated with endothelial dysfunction and subclinical atherosclerosis in subjects 23 Ix JH, Wassel CL, Chertow GM, Koster A, Johnson KC, Tylavsky FA et al. Fetuin-A with nonalcoholic fatty liver disease. Clin Endocrinol (Oxf) 2013; 78:712–717. and change in body composition in older persons. J Clin Endocrinol Metab 2009; 47 Sato M, Kamada Y, Takeda Y, Kida S, Ohara Y, Fujii H et al. Fetuin‐A negatively 94: 4492–4498. correlates with liver and vascular fibrosis in nonalcoholic fatty liver disease sub- 24 Klöting N, Fasshauer M, Dietrich A, Kovacs P, Schön MR, Kern M et al. Insulin- jects. Liver Int; e-pub ahead of print 26 February 2014; doi:10.1111/liv.12478. sensitive obesity. Am J Physiol Endocrinol Metab 2010; 299: E506–E515. 48 Blacher J, Guerin AP, Pannier B, Marchais SJ, London GM. Arterial calcifications, 25 Lavebratt C, Wahlqvist S, Nordfors L, Hoffstedt J, Arner P. AHSG gene variant is arterial stiffness, and cardiovascular risk in end-stage renal disease. Hypertension associated with leanness among Swedish men. Hum Genet 2005; 117:54–60. 2001; 38: 938–942. 26 Thakkinstian A, Chailurkit L, Warodomwichit D, Ratanachaiwong W, Yamwong S, 49 Evrard S, Delanaye P, Kamel S, Cristol J-P, Cavalier E. Vascular calcification: from Chanprasertyothin S et al. Causal relationship between body mass index and pathophysiology to biomarkers. Clin Chim Acta 2014; 438C: 401–414. fetuin‐A level in the asian population: a bidirectional mendelian 50 Mehrotra R, Westenfeld R, Christenson P, Budoff M, Ipp E, Takasu J et al. Serum randomization study. Clin Endocrinol (Oxf) 2013; 81:197–203. fetuin-A in nondialyzed patients with diabetic nephropathy: relationship with 27 Sun Q, Cornelis MC, Manson JE, Hu FB. Plasma levels of fetuin-A and hepatic coronary artery calcification. Kidney Int 2005; 67:1070–1077. enzymes and risk of type 2 diabetes in women in the US. Diabetes 2013; 62: 51 Smith ER, Ford ML, Tomlinson LA, Rajkumar C, McMahon LP, Holt SG. 49–55. Phosphorylated fetuin-A-containing calciprotein particles are associated with 28 Ix JH, Biggs ML, Mukamal KJ, Kizer JR, Zieman SJ, Siscovick DS et al. Association of aortic stiffness and a procalcific milieu in patients with pre-dialysis CKD. Nephrol fetuin-A with incident diabetes mellitus in community-living older adults: the Dial Transplant 2011; 27: 1957–1966. Cardiovascular Health Study. Circulation 2012; 125: 2316–2322. 52 Pasch A, Farese S, Gräber S, Wald J, Richtering W, Floege J et al. Nanoparticle- 29 Kalabay L, Cseh K, Pajor A, Baranyi E, Csakany GM, Melczer Z et al. Correlation of based test measures overall propensity for calcification in serum. J Am Soc Nephrol maternal serum fetuin/alpha2-HS-glycoprotein concentration with maternal 2012; 23: 1744–1752. insulin resistance and anthropometric parameters of neonates in normal 53 Smith ER, Ford ML, Tomlinson LA, Bodenham E, McMahon LP, Farese S et al. pregnancy and gestational diabetes. Eur J Endocrinol 2002; 147: 243–248. Serum calcification propensity predicts all-cause mortality in predialysis CKD. JAm 30 Færch K, Borch-Johnsen K, Holst JJ, Vaag A. Pathophysiology and aetiology of Soc Nephrol 2013; 25: 339–348. impaired fasting glycaemia and impaired glucose tolerance: does it matter 54 Matsui I, Hamano T, Mikami S, Inoue K, Shimomura A, Nagasawa Y et al. Retention for prevention and treatment of type 2 diabetes? Diabetologia 2009; 52: of fetuin-A in renal tubular lumen protects the kidney from nephrocalcinosis 1714–1723. in rats. Am J Physiol Renal Physiol 2013; 304: F751–F760. 31 Stefan N, Sun Q, Fritsche A, Machann J, Schick F, Gerst F et al. Impact of the 55 Lin X, Braymer H, Bray G, York D. Differential expression of insulin receptor tyr- adipokine adiponectin and the hepatokine fetuin-A on the development of type 2 osine kinase inhibitor (fetuin) gene in a model of diet-induced obesity. Life Sci diabetes: Prospective Cohort-and Cross-Sectional Phenotyping Studies. PLoS One 1998; 63: 145–153. 2014; 9: e92238. 56 Samocha-Bonet D, Tam CS, Campbell LV, Heilbronn LK. Raised circulating 32 Andersen G, Burgdorf KS, Sparsø T, Borch-Johnsen K, Jørgensen T, Hansen T et al. fetuin-A after 28-day overfeeding in healthy humans. Diabetes Care 2014; 37: AHSG tag single nucleotide polymorphisms associate with type 2 diabetes and e15–e16. dyslipidemia studies of metabolic traits in 7,683 White Danish subjects. Diabetes 57 Dasgupta S, Bhattacharya S, Biswas A, Majumdar S, Mukhopadhyay S, Ray S. 2008; 57: 1427–1432. NF-kappaB mediates lipid-induced fetuin-A expression in hepatocytes that 33 Siddiq A, Lepretre F, Hercberg S, Froguel P, Gibson F. A synonymous coding impairs adipocyte function effecting insulin resistance. Biochem J 2010; 429: polymorphism in the α2-Heremans-Schmid glycoprotein gene is associated with 451–462. type 2 diabetes in French Caucasians. Diabetes 2005; 54: 2477–2481. 58 Jung TW, Youn B-S, Choi HY, Lee SY, Hong HC, Yang SJ et al. Salsalate and 34 Jensen MK, Bartz TM, Djoussé L, Kizer JR, Zieman SJ, Rimm EB et al. Genetically adiponectin ameliorate hepatic steatosis by inhibition of the hepatokine fetuin-A. elevated fetuin-A levels, fasting glucose levels, and risk of type 2 diabetes: the Biochem Pharmacol 2013; 86:960–969. Cardiovascular Health study. Diabetes Care 2013; 36: 3121–3127. 59 Weyer C, Funahashi T, Tanaka S, Hotta K, Matsuzawa Y, Pratley RE et al. Hypoa- 35 Stefan N, Häring H-U. Circulating fetuin-A and free fatty acids interact to predict diponectinemia in obesity and type 2 diabetes: close association with insulin insulin resistance in humans. Nat Med 2013; 19:394–395. resistance and hyperinsulinemia. J Clin Endocrinol Metab 2001; 86: 1930–1935. 36 Jensen MK, Bartz TM, Mukamal KJ, Djoussé L, Kizer JR, Tracy RP et al. Fetuin-A, 60 Hussey SE, Lum H, Alvarez A, Cipriani Y, Garduño-Garcia J, Anaya L et al. type 2 diabetes, and risk of cardiovascular disease in older adults: the Cardio- A sustained increase in plasma NEFA upregulates the Toll-like receptor network in vascular Health study. Diabetes Care 2013; 36:1222–1228. human muscle. Diabetologia 2013; 57:582–591. 37 Laughlin GA, Cummins KM, Wassel CL, Daniels LB, Ix JH. The association of 61 Takata H, Ikeda Y, Suehiro T, Ishibashi A, Inoue M, Kumon Y et al. High glucose fetuin-A with cardiovascular disease mortality in older community-dwelling induces transactivation of the alpha2-HS glycoprotein gene through the ERK1/2 adults: the Rancho Bernardo study. J Am Coll Cardiol 2012; 59: 1688–1696. signaling pathway. J Atheroscler Thromb 2009; 16:448–456. 38 Lorant DP, Grujicic M, Hoebaus C, Brix J-M, Hoellerl F, Schernthaner G et al. 62 Ou H-Y, Wu H-T, Hung H-C, Yang Y-C, Wu J-S, Chang C-J. Endoplasmic reticulum Fetuin-A levels are increased in patients with type 2 diabetes and peripheral stress induces the expression of fetuin-A to develop insulin resistance. arterial disease. Diabetes Care 2011; 34: 156–161. Endocrinology 2012; 153: 2974–2984. 39 Eraso LH, Ginwala N, Qasim AN, Mehta NN, Dlugash R, Kapoor S et al. Association 63 Chatterjee P, Seal S, Mukherjee S, Kundu R, Mukherjee S, Ray S et al. of lower plasma fetuin-a levels with peripheral arterial disease in type 2 diabetes. Adipocyte fetuin-a contributes to macrophage migration into adipose tissue and Diabetes Care 2010; 33:408–410. polarization of macrophages. J Biol Chem 2013; 288: 28324–28330. 40 Stefan N, Weikert C, Ix JH, Fritsche A, Häring H-U. Association of lower plasma 64 Srinivas P, Wagner AS, Reddy LV, Deutsch D, Leon MA, Goustin AS et al. Serum fetuin-A levels with peripheral arterial disease in type 2 diabetes response to alpha 2-HS-glycoprotein is an inhibitor of the human insulin receptor at the Eraso et al. Diabetes Care 2010; 33:e55–e55. tyrosine kinase level. Mol Endocrinol 1993; 7: 1445–1455.

International Journal of Obesity (2015) 734 – 741 © 2015 Macmillan Publishers Limited Fetuin-A in obesity and related complications JF Trepanowski et al 741 65 Auberger P, Falquerho L, Contreres JO, Pages G, Cam GL, Rossi B et al. 77 Lavebratt C, Dungner E, Hoffstedt J. Polymorphism of the AHSG gene is associated Characterization of a natural inhibitor of the insulin receptor tyrosine kinase: with increased adipocyte β2-adrenoceptor function. J Lipid Res 2005; 46: cDNA cloning, purification, and anti-mitogenic activity. Cell 1989; 58: 2278–2281. 631–640. 78 Blüher M, Rudich A, Klöting N, Golan R, Henkin Y, Rubin E et al. Two patterns of 66 Rauth G, Pöschke O, Fink E, Eulitz M, Tippmer S, Kellerer M et al. The nucleotide adipokine and other biomarker dynamics in a long-term weight loss intervention. and partial amino acid sequences of rat fetuin. Eur J Biochem 1992; 204: Diabetes Care 2012; 35:342–349. 523–529. 79 Reinehr T, Roth CL. Fetuin-A and its relation to metabolic syndrome and fatty liver 67 Mathews ST, Srinivas PR, Leon MA, Grunberger G. Bovine fetuin is an inhibitor of disease in obese children before and after weight loss. J Clin Endocrinol Metab insulin receptor tyrosine kinase. Life Sci 1997; 61: 1583–1592. 2008; 93:4479–4485. 68 Mathews ST, Rakhade S, Zhou X, Parker GC, Coscina DV, Grunberger G. Fetuin-null 80 Jüllig M, Yip S, Xu A, Smith G, Middleditch M, Booth M et al. Lower fetuin-A, retinol mice are protected against obesity and insulin resistance associated with aging. binding protein 4 and several metabolites after gastric bypass compared to Biochem Biophys Res Commun 2006; 350:437–443. sleeve gastrectomy in patients with type 2 diabetes. PLoS One 2014; 9: e96489. 69 Mathews ST, Singh GP, Ranalletta M, Cintron VJ, Qiang X, Goustin AS et al. 81 Schultes B, Frick J, Ernst B, Stefan N, Fritsche A. The effect of 6-weeks of aerobic Improved insulin sensitivity and resistance to weight gain in mice null for the exercise training on serum fetuin-A levels in non-diabetic obese women. Exp Clin Ahsg gene. Diabetes 2002; 51: 2450–2458. Endocrinol Diabetes 2010; 118: 754–756. 70 Srinivas PR, Deutsch DD, Mathews ST, Goustin AS, Leon MA, Grunberger G. 82 Yang SJ, Hong HC, Choi HY, Yoo HJ, Cho GJ, Hwang TG et al. Effects of a three‐ Recombinant human alpha 2-HS glycoprotein inhibits insulin-stimulated mito- month combined exercise programme on fibroblast growth factor 21 and fetuin‐ genic pathway without affecting metabolic signalling in Chinese hamster ovary A levels and arterial stiffness in obese women. Clin Endocrinol (Oxf) 2011; 75: cells overexpressing the human insulin receptor. Cell Signal 1996; 8:567–573. 464–469. 71 Chen H, Srinivas PR, Cong L-N, Li Y, Grunberger G, Quon MJ. α2-Heremans schmid 83 Malin SK, Mulya A, Fealy CE, Haus JM, Pagadala MR, Scelsi AR et al. Fetuin-A is glycoprotein inhibits insulin-stimulated Elk-1 phosphorylation, but not glucose linked to improved glucose tolerance after short-term exercise training in non- transport, in rat adipose cells 1. Endocrinology 1998; 139: 4147–4154. alcoholic fatty liver disease. J Appl Physiol 2013; 115:988–994. 72 Goustin AS, Derar N, Abou-Samra AB. Ahsg-fetuin blocks the metabolic arm of 84 Xu XJ, Valentine RJ, Ruderman NB. AMP-activated protein kinase (AMPK): does this insulin action through its interaction with the 95-kD β-subunit of the insulin master regulator of cellular energy state distinguish insulin sensitive from insulin receptor. Cell Signal 2013; 25: 981–988. resistant obesity? Curr Obes Rep 2014; 3:248–255. 73 Pal D, Dasgupta S, Kundu R, Maitra S, Das G, Mukhopadhyay S et al. Fetuin-A acts 85 Le Cam A, Magnaldo I, Le Cam G, Auberger P. Secretion of a major phosphory- as an endogenous ligand of TLR4 to promote lipid-induced insulin resistance. Nat lated glycoprotein by hepatocytes. Characterization of specific antibodies and Med 2012; 18: 1279–1285. investigations of the processing, excretion kinetics, and phosphorylation. J Biol 74 Yang L, Zhang B, Song J, Miura K, Seki E. 695 Fetuin-a binding to TLR4 regulates Chem 1985; 260: 15965–15971. NASH and fibrosis: the role of TRIF. Gastroenterology 2014; 146: S-922. 86 Mathews S, Ren G, He X, Bowers R, Araya-Ramirez F, Littlefield L et al. Plasma 75 Lumeng CN, Bodzin JL, Saltiel AR. Obesity induces a phenotypic switch in adipose fetuin-A and phosphofetuin-A (Ser312) responses to a single or short-term tissue macrophage polarization. J Clin Invest 2007; 117:175–184. repeated bout of exercise in obese and normal-weight individuals (1028.2). FASEB 76 Dahlman I, Eriksson P, Kaaman M, Jiao H, Lindgren C, Kere J et al. α2-Heremans– J 2014; 28: 1028.2. Schmid glycoprotein gene polymorphisms are associated with adipocyte 87 Yin J, Peng Y, Wu J, Wang Y, Yao L. Toll-like receptor 2/4 links to free fatty acid- insulin action. Diabetologia 2004; 47: 1974–1979. induced inflammation and β-cell dysfunction. J Leukoc Biol 2014; 95:47–52.